In-Vehicle Intelligent Agents in Fully Autonomous Driving: The Effects of Speech Style and Embodiment Together and Separately

Abstract

Speech style and embodiment are two widely researched characteristics of in-vehicle intelligent agents (IVIAs). This study aimed to investigate the influence of speech style (informative vs. conversational) and embodiment (voice-only vs. robot) and their interaction effects on driver-agent interaction. We conducted a driving simulator experiment, where 24 young drivers experienced four different fully autonomous driving scenarios, accompanied by four types of agents each, and completed subjective questionnaires about their perception towards the agents. Results showed that both conversational agents and robot agents promoted drivers' likability and perceived warmth. These two features also demonstrated independent impacts. Conversational agents received higher anthropomorphism and animacy scores, while robot agents received higher competence and lower perceived workload scores. The pupillometry indicated that drivers were more engaged while accompanied by conversational agents. Our findings are able to provide insights on applying different features to IVIAs to fulfill various user needs in highly intelligent autonomous vehicles.